Mahalanobis classification system (MCS) integrated with binary particle swarm optimization for robust quality classification of complex metallic turbine blades

Performing non-destructive testing on metallic components with very complex geometries, such as turbine blades, is very challenging. To inspect such components, powerful and robust non-destructive inspection protocols must be defined. Process Compensated Resonance Testing (PCRT) is a relatively novel approach that records a broadband vibrational fingerprint for each component and employs vibrational features such as resonant frequency, quality factor, and amplitude. These features are used in the Mahalanobis Taguchi System to classify the parts in terms of their quality, i.e. Good/Bad.

In the present study, a two-stage MCS classification approach, coupled with Binary Particle Swarm Optimization, is proposed to optimize the process of selecting the most significant features and to search for the optimal decision boundary to discriminate healthy and unhealthy components. Further, the proposed MCS enables the features to be mapped into a higher dimensional Mahalanobis Distance space, thereby enhancing the performance of classification. An experimental case study on equiaxed Nickel alloy first-stage turbine blades, with very complex geometry and various damages, demonstrates the high classification accuracy and robustness of the developed MCS approach.